Squirty Design Notes P. McGill 7Jul14 * Housing is left-over Ti BIOSA housing with 2.36” ID. With a battery of 9 alkaline C cells, there is about 3” of length left in the housing, not counting end caps, connectors, etc. * According to Seabird 5T data sheet (http://www.seabird.com/pdf_documents/Datasheets/05T-05PbrochureMay13.pdf) the pump draws 640 mA at 12 V. * With 9 alkaline cells, starting battery voltage is over 13.5 V, and ending voltage at 0.8 V/cell is 7.2 V. Make power supply output 12 V at up to 1 A over input range from 7 to 15 V. * Buck-boost converter over 7 to 15 V is inefficient and difficult to design. Could instead get pump rated to run at 6 V and make a buck-only converter. - Ended up using a LM2936 linear regulator to power the PIC and Okuda’s “MRS BRS Pump Regulator” board to limit voltage to the pump. Not super efficient, but battery has plenty of energy to power system for a year. * Circuit board should be 2” diameter to allow wires to pass around edges. - Instead used Craig’s postage-stamp-sized board with some anti-static bubble wrap around it. * Circuit should have push-button switch to activate pump for testing, but switch must be mounted in such a way as to prevent inadvertent activation when housing is closed. - Instead programmed software to run pump for 10 sec after first power-up and then at programmed interval thereafter. * Can use Craig Okuda’s “PIC Stirrer Controller” (PSC) board. PSC has high-side switch to feed pump from raw battery voltage, which is fine for our application. Switch is controlled by PIC10F202 micro. * PSC draws 450 uA, which is too much for us. * PSC uses a non-standard ICSP connector pinout: 1 - Vdd 2 - PGD 3 - PGC 4 - MCLR*/Vpp 5 - GND * Replaced PSC’s PIC10F202 with PIC10LF322 and LT1129CST-3.3 with LM2936-3.3. PSC now draws 20 uA. Ran test with power meter at 13.7 V, circuit uses 0.424 mW. In a year this would be 8760 h/yr x 0.424 mW / 1000 mW/W = 3.7 Wh/yr. This is about 1/3 of an alkaline C cell. This doesn’t include the power to run the pump.